FR3044504A1 - IMAGE CAPTURE DEVICE AND CONDUCTOR MONITORING DEVICE USING SUCH IMAGE CAPTURE DEVICE - Google Patents

Abstract

An image capture device comprises: - an image acquisition module (IMG) with an adjustable acquisition duration (T); - a determination module (20) of a central tendency indicator (M) relative to the luminance values associated respectively with the pixels of at least one region (R) of the image (IMG); a module (30) for adjusting the acquisition duration (T) according to the determined central tendency indicator (M). A driver monitoring device including such a capture device is also described.

Description

Technical field to which the invention relates

The present invention relates to the acquisition of images and their use, for example for monitoring a vehicle driver.

It relates more particularly to an image capture device and a device for monitoring a conductor using such an image capture device. The invention is particularly advantageous in the case where it is desired to determine, in a repeatable manner, a level of distraction and / or a level of somnolence of the driver.

Technological background

For example, in vehicle driver monitoring devices, an image pickup device comprising an image acquisition module with an adjustable acquisition duration is used.

When the images are then used in a subsequent processing process, for example, as part of the aforementioned monitoring devices, in order to determine a level of distraction and / or a level of driver sleepiness, it is naturally desirable to obtain repeatable results, regardless of contextual changes that may affect the acquired image.

Such changes may be related for example to variations in illumination of the captured scene, as is the case when the distance between an illumination device and an object illuminated by it varies.

Object of the invention

In this context, the present invention proposes an image capture device comprising an image acquisition module with an adjustable acquisition duration, characterized by a module for determining a central tendency indicator relative to the values of the image. luminance associated respectively with the pixels of at least one region of the image and with a module for adjusting the acquisition duration according to the determined central trend indicator.

The duration of acquisition, and consequently the luminance of the pixels of the images acquired, can thus be adjusted by taking into account a general level of luminance in the aforementioned region, for example in order to obtain a predetermined level of luminance, which improves the repeatability of subsequent treatments.

The adjustment module is for example designed to adjust the acquisition time according to the difference between the determined central trend indicator and a set value, possibly so that said central tendency indicator converges towards a setpoint predetermined. To do this, a derivative integral proportional (PID) type regulation mechanism can be used, as described later. The central tendency indicator relating to the luminance values is, for example, the average of the luminance values (values which are, as indicated above, respectively associated with the pixels of the aforementioned region of the image). However, another measure of central tendency, such as the median of these luminance values, could alternatively be used.

It can further be provided that the image capture device comprises an infrared illumination module of an environment facing the acquisition module. A control module can then activate the illumination module in synchronism with the acquisition module. In addition, the acquisition module may be sensitive to infrared radiation. The use of a system working in the infrared makes it possible to obtain images (in the example described below, images of the driver's face of the vehicle) even during night-time use, without however disturbing the environment (namely here without dazzling the driver).

The image capture device may include a face detection block in the acquired image and / or an identification block of said region within the detected face. The luminance of the pixels is thus regulated on a region of interest, or on a region close to a region of interest, for the subsequent processing of the image (here in the context of monitoring the driver). The invention also proposes a device for monitoring a vehicle driver comprising an image capture device as proposed above.

Such a monitoring device comprises an acquired image processing module, designed for example to provide information indicative of a level of unfitness to driving. As described below, such a level of incapacity can be, for example, a level of somnolence and / or a level of distraction, determined on the basis of characteristics extracted from the acquired image (or a sequence of images thus acquired). ).

Detailed description of an example of realization

The following description with reference to the accompanying drawings, given as non-limiting examples, will make it clear what the invention consists of and how it can be achieved.

In the accompanying drawings: FIG. 1 diagrammatically represents an example of a vehicle driver monitoring device in accordance with the invention; FIG. 2 is a logic diagram illustrating an exemplary mechanism for regulating the duration of acquisition of an image; and FIG. 3 represents an exemplary region identified in the acquired image.

Figure 1 shows schematically the main elements of a system in which the invention can be implemented.

This system comprises a monitoring device 2 of a vehicle driver (here a motor vehicle). This monitoring device 2 is embedded in the vehicle, for example in a cabin H of the vehicle, possibly in a region close to a windshield W of the vehicle.

The monitoring device 2 comprises an image acquisition module 4 (for example a camera, here an infrared camera), an illumination module 6 (for example an infrared radiation emitter) and a computer 8. These elements can be united in a single dedicated box; alternatively, the computer may be located remote from the image acquisition module and / or the illumination module and can then exchange data with these modules for example via an on-board network.

The image acquisition module 4 is designed to generate an image of the environment facing it. Such an image is formed of a matrix of pixels. Each pixel is associated with a luminance value corresponding to the intensity of the radiation (here infrared) received, during an adjustable acquisition time T, at the level of the image acquisition module 4 coming from a particular direction in the environment facing the image acquisition module 4.

The acquisition module 4 is sensitive to a frequency range of the radiation (here in the infrared) which covers at least in part the frequency range of the radiation emitted by the illumination module 6.

The acquisition module 4 comprises, for example, an image sensor and an optical system returning to the image sensor an image of a part of the environment facing the acquisition module 4.

The acquisition module 4 is arranged in the vehicle so as to be oriented in the direction of the driver's head C of the vehicle so that the image acquired by the acquisition module 4 includes in particular the face V of the driver C.

One could alternatively use another image acquisition module (for example another infrared camera) in order to build a stereoscopic image of the driver's head.

The computer 8 receives the IMG acquired image by the acquisition module 4, that is to say in practice the luminance values respectively associated with the pixels of the IMG acquired image.

The computer 8 then proceeds to the analysis of the acquired image IMG, or of a sequence of images thus acquired, in order to identify the driver's face V and / or certain areas of the driver's face V and to determine a level of distraction ND of the driver C and a level of somnolence Ns of the driver C.

The computer 8 determines, for example, the level of somnolence Ns of the conductor C by identifying the areas of the IMG image corresponding to the eyes of the driver C and by measuring the duration and / or the frequency of blinking of the eyes of the driver C.

The computer 8 determines, for example, the level of distraction ND of the driver C by determining the posture of the driver's head C and the evolution of this posture over time.

The calculator 8 can also evaluate (by analysis of the IMG acquired image or of a sequence of acquired images) and use, in order to determine the level of distraction ND and / or the level of somnolence Ns, the direction of gaze of the driver. C or the evolution of this direction of gaze over time.

The calculator 8 can also evaluate (by analysis of the acquired image IMG or of a sequence of acquired images) and use, to determine the level of distraction ND and / or the level of sleepiness Ns, the diameter of the pupil of at least one eye of the conductor C (and precisely the variations of this diameter).

The computer 8 is also designed to adjust the acquisition time T of the acquisition module 4 according to a central trend indicator of the luminance distribution in at least one region of an image acquired by the acquisition module 4, for example according to the process described below with reference to FIG.

For example, the computer 8 controls the opening of a shutter of the acquisition module 4 and then, after a duration equal to the acquisition time T, the closure of the shutter. The computer 8 may optionally further control in synchronism (that is to say during the duration of opening of the flap) the activation of the illumination module 6 (the illumination module being on the other hand inactive when the flap is closed).

FIG. 2 is a logic diagram illustrating an exemplary mechanism for regulating the aforementioned acquisition duration T, implemented here within the computer 8 as already indicated.

This mechanism is described here in terms of functional modules. When the computer 8 is made on the basis of a processor (for example a microprocessor), each module is implemented by computer program instructions (stored for example in the computer 8) designed to perform the task. function of the module concerned when these instructions are executed by the processor.

It is understood that each module also corresponds to a step of a method for determining the acquisition time T.

As already indicated above, a control module 10 controls the activation of the illumination module 6 and the acquisition of an IMG image by the acquisition module 4 during an acquisition period T.

During the initialization of the regulation mechanism (that is to say during the first implementation of the acquisition step of an IMG image by the acquisition module 4), the acquisition duration T is for example fixed at a predetermined value (stored for example in the computer 8). Subsequently, the acquisition time T is determined by the adjustment module 30 described below. The IMG acquired image (i.e., in practice, the luminance values of the pixels of the image) is transmitted to a module for determining a central tendency indicator 20.

This determination module 20 is designed to determine a central tendency indicator M relative to the set of luminance values associated respectively with the pixels of a region R of the IMG image. The central tendency indicator M used here is the average of the luminance values associated respectively with the pixels of said region R. Alternatively, another measure of central tendency, such as the median, could be used.

The determination module 20 comprises an identification sub-module 22 of the region R in the IMG image and a calculation sub-module 24 of the central tendency indicator M of the luminance of the pixels of the region R.

The identification sub-module 22 is for example designed to identify, by means of a shape recognition algorithm, the face V of the driver present in the acquired image IMG, then to locate the region R in the image acquired. IMG relative to the identified face V.

FIG. 3 shows an example in which the region R is a region of the front of the conductor, located with respect to the face V of the driver identified in the acquired image IMG by means of a shape recognition algorithm.

Note that, alternatively, several (possibly unrelated) regions could be identified by the identification sub-module 22. The IMG image is also used (possibly within a sequence of images) within a processing module 50 for evaluating a level of distraction Nd and / or a level of sleepiness Ns of the conductor C, as already described above.

The calculation sub-module 24 determines the central tendency indicator M (here, as already indicated, the average) luminance values respectively associated with all the pixels of the region R.

When several regions are identified by the identification sub-module 22, the calculation sub-module 24 determines for example the central tendency indicator on all the identified regions. The central tendency indicator M determined by the determination module 20 is transmitted to an adjustment module 30 of the acquisition duration T according to this central tendency indicator M.

The adjustment module 30 uses here a PID type control mechanism (for "Proportional Integral Derivative").

The adjustment module 30 comprises an evaluation sub-module 32 of a difference ε to the setpoint, a regulation sub-module 34 and a correction sub-module 36 of the (current) acquisition duration T.

The error evaluation sub-module 32 calculates the difference (or difference) ε between the central trend indicator M received from the determination module 20 and a predetermined setpoint MCons (stored for example in the computer 8).

The difference ε thus calculated is transmitted to the regulation submodule 34 which calculates on this basis a variation of duration At, as already indicated in accordance with a PID type control mechanism. However, it would be possible alternatively to use another regulation mechanism.

The variation of duration Δt calculated by the regulation sub-module 34 is transmitted to the correction sub-module 36 which modifies the duration of acquisition T by adding to it the variation of duration At. be negative and that the acquisition duration T is then reduced.)

The modified (or corrected) acquisition time T can thus be transmitted to the control module 10 for use during the next acquisition of an IMG image.

The acquisition duration T is thus regulated so as to obtain, in the region R of the images acquired IMG, luminance values whose central tendency indicator (here the average) corresponds (possibly after several iterations of the process described above). above) to the Mcons-setpoint. This improves the repeatability of the treatments performed on the acquired IMG images, here to determine a level of sleepiness Ns of the driver C and / or a level of distraction ND of the driver C.

Claims (12)

An image capture device comprising: an acquisition module (4) of an image (IMG) with an adjustable acquisition duration (T); characterized by: - a determination module (20) of a central tendency indicator (M) relative to the luminance values associated respectively with the pixels of at least one region (R) of the image (IMG); a module (30) for adjusting the acquisition duration (T) according to the determined central tendency indicator (M). An image capture device according to claim 1, wherein the adjustment module (30) is adapted to adjust the acquisition time (T) as a function of the difference (ε) between the central tendency indicator. (M) determined and a set value (MCons) · An image capture apparatus according to claim 1 or 2, wherein the central tendency indicator (M) relating to the luminance values is the average of the luminance values. 4. An image capture device according to one of claims 1 to 3, comprising an infrared illumination module (6) of an environment facing the acquisition module (4). An image capture device according to claim 4, comprising a control module (10) adapted to activate the illumination module (6) in synchronism with the acquisition module (4). 6. An image capture device according to one of claims 1 to 5, wherein the acquisition module (4) is sensitive to infrared radiation. An image capture device according to one of claims 1 to 6, comprising a face detection block (V) in the acquired image and an identification block of said region (R) within the face (V) detected. 8. A vehicle driver monitoring device comprising an image pickup device according to one of claims 1 to 7. 9. Monitoring device according to claim 8, comprising a processing module (50) of the acquired image (IMG). 10. Monitoring device according to claim 9, wherein the processing module (50) is designed to deliver information (Ns, ND) indicative of a level of unfitness to driving. 11. Monitoring device according to claim 10, wherein the level of unfitness is a level of sleepiness (Ns) · 12. Monitoring device according to claim 10, wherein the level of unfitness is a level of distraction (Nd).